Brain Changes in Responders vs. Non-Responders in Chronic Migraine: Markers of Disease Reversal

Abstract

The aim of this study was to identify structural and functional brain changes that accompanied the transition from chronic (CM; ≥15 headache days/month) to episodic (EM; <15 headache days/month) migraine following prophylactic treatment with onabotulinumtoxinA (BoNT-A). Specifically, we examined whether CM patients responsive to prophylaxis (responders; n = 11), as evidenced by a reversal in disease status (defined by at least a 50% reduction in migraine frequency and <15 headache days/month), compared to CM patients whose migraine frequency remained unchanged (non-responders; n = 12), showed differences in cortical thickness using surface-based morphometry. We also investigated whether areas showing group differences in cortical thickness displayed altered resting-state functional connectivity (RS-FC) using seed-to-voxel analyses. Migraine characteristics measured across groups included disease duration, pain intensity and headache frequency. Patient reports of headache frequency over the 4 weeks prior to (pre-treatment) and following (post-treatment) prophylaxis were compared (post minus pre) and this measure served as the clinical endpoint that determined group assignment. All patients were scanned within 2 weeks of the post-treatment visit. Results revealed that responders showed significant cortical thickening in the right primary somatosensory cortex (SI) and anterior insula (aINS), and left superior temporal gyrus (STG) and pars opercularis (ParsOp) compared to non-responders. In addition, disease duration was negatively correlated with cortical thickness in fronto-parietal and temporo-occipital regions in responders but not non-responders, with the exception of the primary motor cortex (MI) that showed the opposite pattern; disease duration was positively associated with MI cortical thickness in responders versus non-responders. Our seed-based RS-FC analyses revealed anti-correlations between the SI seed and lateral occipital (LOC) and dorsomedial prefrontal cortices (DMPFC) in responders, whereas non-responders showed increased connectivity between the ParsOp seed and LOC. Overall, our findings revealed distinct morphometric and functional brain changes in CM patients that reverted to EM following prophylactic treatment compared to CM patients that showed no change in disease status. Elucidating the CNS changes involved in disease reversal may be critical to discovering interventions that prevent or slow the progression of CM. Such changes may aid in the evaluation of treatments as well as provide markers for disease "de-chronification".

Keywords: BOTOX®; fMRI; gray matter; headache; migraine transformation; network connectivity; pain; preventative therapy.

Figure 1

Statistical group maps displaying differences in cortical thickness in treatment responders relative to non-responders rendered onto an average study-specific inflated cortical surface for the right and left hemisphere. Red clusters correspond to areas displaying greater cortical thickness in responders vs. non-responders for the primary somatosensory cortex (SI) and anterior insula (aINS) in the right hemisphere and the pars opercularis (ParsOp) and superior temporal gyrus (STG) in the left hemisphere. Bottom right brain is angled to illustrate the cluster in the left STG.

Figure 2

Statistical group maps rendered onto an average study-specific inflated surface for the left and right hemispheres displaying areas showing significant group differences in slopes for associations between cortical thickness and disease duration while controlling for age. Scatter plots illustrate differences in slopes for each group while controlling for age (blue filled circles and line correspond to treatment responders and red filled circles and line correspond to non-responders). Abbreviations: MI, Primary motor cortex; DLPFC, dorsolateral prefrontal cortex; LOC, lateral occipital cortex; ITG, inferior temporal gyrus; PPC, posterior parietal cortex; PCC, posterior cingulate cortex.

Figure 3

Statistical group maps rendered onto an inflated cortical surface for the seed in the right primary somatosensory cortex (SI) showed greater anti-correlated functional connectivity with clustered voxels in the lateral occipital cortex (LOC) and dorsomedial prefrontal cortex (DMPFC) for responders vs. non-responders.

Figure 4

Statistical group maps rendered onto an inflated cortical surface for the left parsopercularis (ParsOp) seed showing greater positive connectivity for voxels in the lateral occipital cortex (LOC) in responders compared to non-responders.